Ch6: The

Introduction

Review: A continuous can assume any value between two endpoints.

Many continuous random variables have an approximately normal distribution, which we can use the distribution of a normal random variable to make about the variable.

CH6: The Normal Distribution Santorico - Page 175 Example: Heights of randomly selected women

CH6: The Normal Distribution Santorico - Page 176 Section 6-1: Properties of a Normal Distribution

A normal distribution is a continuous, symmetric, bell-shaped distribution of a variable.

The theoretical shape of a normal distribution is given by the

mathematical formula (x)2  e 2 2 y  ,  2

where  and  are the and standard deviations of the , respectively.

 Review: The and are parameters and hence describe the  population .

CH6: The Normal Distribution Santorico - Page 177 The shape of a normal distribution is fully characterized by its mean  and standard .

  specifies the location of the distribution.    specifies the spread/shape of the distribution.

 

CH6: The Normal Distribution Santorico - Page 178

CH6: The Normal Distribution Santorico - Page 179 Properties of the Theoretical Normal Distribution

 A normal distribution curve is bell-shaped.  The mean, , and are equal and are located at the center of the distribution.  A normal distribution curve is unimodal.  The curve is symmetric about the mean.  The curve is continuous (no gaps or holes).  The curve never touches the x-axis (just approaches it).  The total area under a normal distribution curve is equal to 1.0 or 100%.  Review: The Empirical Rule applies (68%-95%-99.7%)

CH6: The Normal Distribution Santorico - Page 180

CH6: The Normal Distribution Santorico - Page 181 The Standard Normal Distribution

Since each normally distributed variable has its own mean and , the shape and location of these curves will vary.

To simplify making statistical inference using the normal distribution, we use the standard normal distribution.

Standard normal distribution - a normal distribution with mean 0 and a standard deviation of 1.

CH6: The Normal Distribution Santorico - Page 182 All normally distributed variables can be transformed into a standard normally distributed variable using the formula for the standard (z-score):

valuemean X  z  or z  . standard deviation 

The z-score for an observation is the number of standard deviations the observation lies from the mean.   The letter Z is used to denote a standard normal random variable.



CH6: The Normal Distribution Santorico - Page 183 Example: The mean resting pulse rates for men are normally distributed with a mean of 70 and a standard deviation of 8. A man has a pulse of 80. Find the corresponding z-score.

We know: =70, =8, X=80

X  80 70 10 z     1.25  88

The man’s pulse is ______standard deviations above the mean because ______is positive.

CH6: The Normal Distribution Santorico - Page 184 Example: The average speed of drivers on I-25 is 63 mph with a standard deviation of 5 mph. Assuming the data are normally distributed, find the z-score for the following speeds clocked by police officers:

74 mph z=

51 mph z=

82 mph z=

CH6: The Normal Distribution Santorico - Page 185 Finding Areas/Probabilities for the Standard Normal Distribution

Table E in Appendix C gives the area/probability under the standard normal distribution curve to the left of z-values between -3.49 and 3.49 in increments of .01.

 To find P (Z  z) (the probability that a standard normal random variable Z is less than or equal to the value z ), we match our value of z with the left column and top row of Table E (by adding the numbers together). The row and column where these numbers intersect gives us the   probability or area under the standard normal curve to the left of z.  Note: Use .0001 when z  3.50 and .9999 when z  3.50.

CH6: The Normal Distribution Santorico - Page 186   Examples:

P (Z  3.49) .0002

P (Z  1.31) .0951

 P (Z 0) .5000

 P (Z 1.96) .9750

 P (Z 3.49) .9998

 P (Z 1.43) 

 P (Z  1.82)   CH6: The Normal Distribution Santorico - Page 187  Finding the Area Under the Standard Normal Distribution Curve

1. To the left of any z-value: look up the z value in the table and use the area given.

CH6: The Normal Distribution Santorico - Page 188 Examples:

Find the area to the left of z  0.84, i.e., find P (Z 0.84).

  Find the area to the left of z = -1.32, i.e., find P (Z  1.32).



CH6: The Normal Distribution Santorico - Page 189 2. To the right of any z-value: look up the z value and subtract the area from 1.

CH6: The Normal Distribution Santorico - Page 190 Examples:

Find the area to the right of z  0.21, i.e., find P (Z  0.21).

P(Z > -0.21) = 1 – P(Z  -0.21) = 1 – 0.4168 = 0.5832

 

Find the area to the right of z 1.52, i.e., find P (Z 1.52).

 

CH6: The Normal Distribution Santorico - Page 191 3. Between any two z-values: look up both z values and subtract the smaller area from the larger area.

CH6: The Normal Distribution Santorico - Page 192 Examples:

Find the area between z  1.71 and z  0.45, i.e., find

PZPZPZ( 1.71   0.45)  (  0.45)  (   1.71)

 0.6736 - 0.0436  0.63

Find the area between z=1.43 and z=3.01, i.e., find P (1.43 Z 3.01).



CH6: The Normal Distribution Santorico - Page 193 Note: For a continuous random variable X (not a discrete random variable), the probability that X equals a specific number is 0.  Let c be some number (like 1 or 7). Then P (X  c) 0.

Intuitive explanation:  If we had ten numbers, 1, 2, …, 10, and X is the number we  see, P (X 2) 1/10.  If we had one thousand numbers, 1, 2, …, 1000, and X is the number we see, P (X 2) 1/1000.  For a continuous random variable, there are an infinite  number of possibilities, so P (X  c) 1/ 0. (This is bad mathematics, but it gets the point across). 



CH6: The Normal Distribution Santorico - Page 194 Thus, for a continuous random variable:

P (Z  z)  P(Z  z) and

P (Z  z)  P(Z  z) and

 P (z1  Z  z2 )  P(z1  Z  z2 )  P(z1  Z  z2 )  P(z1  Z  z2 ) 



CH6: The Normal Distribution Santorico - Page 195 Section 6-2: Applications of the Normal Distribution

Finding Probabilities for Normal Distributions

To find probabilities for any normal distribution with mean  and standard deviation , we simply transform the values on the original scale (call these x, x 1, or x 2) using the formula

valuemean x    z  or z  , standard deviation    and then use the techniques from the previous section.

 

CH6: The Normal Distribution Santorico - Page 196 Example: An American household generates an average of 28 pounds per month of newspaper for garbage, with a standard deviation of 2 pounds. Assume the amount of newspaper garbage is normally distributed. If a household is selected at random, find the probability of generating between 27 and 31 pounds of newspaper garbage per month.

X = garbage. We want P(27 < X < 31).

Turn into a Z score so we can use the standard normal distribution.

27X   31     27  28 31  28  PPZ            22  PZ(  0.5   1.5) X   PZPZ(  1.5)  (   0.5) Z   and 0.9332 0.3085  0.6247

CH6: The Normal Distribution Santorico - Page 197 Example: College women’s heights are normally distributed with  65 inches and   2.7 inches. What is the probability that a randomly selected college woman is 62 inches or shorter?   XX 65 Z  X=height and  2.7 .

X 62 PXP( 62)    62 65 PZ 2.7 We want PZ( -1.11)  0.1335

CH6: The Normal Distribution Santorico - Page 198 Example: What is the probability that a randomly selected college woman is between 63 and 69 inches tall?

CH6: The Normal Distribution Santorico - Page 199 Sometimes we need to find the value X that corresponds to a certain . (e.g., what value corresponds to the 98th percentile?)

To solve this problem:

1. Find the z that corresponds to this percentile for the standard normal distribution. a. The area to the left of this z-score has the area closest to this percentile.

2. X    z .



CH6: The Normal Distribution Santorico - Page 200 Example: Mensa is a society of high-IQ people whose members have a score on an IQ test at the 98th percentile or higher. The mean IQ score is 100 with a standard deviation of 16. What IQ score do you need to get into Mensa?

1. First determine what z-score is related to the 98th percentile (to be in the 98th percentile means that 98% of IQ test takers scored at or below your score) (i.e. 0.980).

Z = 2.06

2. What is the IQ for the 98th percentile using the formula for X given above?

Z = +Z = 100 + 2.06(16) = 132.96

CH6: The Normal Distribution Santorico - Page 201 Example: College women’s heights are normally distributed with a  65 and   2.7. What is the 54th percentile of this population?

th 1. Find Z-score for 54 percentile: 

2. Find corresponding value for X:

CH6: The Normal Distribution Santorico - Page 202 Example: An American household generates an average of 28 pounds per month of newspaper for garbage with a standard deviation of 2 pounds. Assume the amount of newspaper garbage is normally distributed. What is the 34th percentile for the amount of newspaper garbage American households produce per month?

CH6: The Normal Distribution Santorico - Page 203 Determining Normality

The easiest way to determine if a distribution is bell-shaped is to draw a for the data and check its shape. If the histogram is NOT approximately bell-shaped then the distribution is NOT normally distributed.

The of a data set can be checked by using the Pearson’s index, PI, of skewness.

3(X median) PI  s

If the index is greater than or equal to +1 or less than or equal to -1, it can be concluded that the data are significantly skewed. 

CH6: The Normal Distribution Santorico - Page 204 If the index is smaller than -1, then the distribution is negatively skewed. If the index is greater than +1, then the distribution is positively skewed.

The data should also be checked for outliers since they can have a big effect on normality.

If the histogram is bell-shaped, the PI is between -1 and +1, and there are no outliers then it can be concluded that the distribution is normally distributed.

CH6: The Normal Distribution Santorico - Page 205 Example: A survey of 18 high-technology firms showed the number of days’ inventory they had on hand. Determine if the data are normally distributed.

CH6: The Normal Distribution Santorico - Page 206 For these data, the mean is 79.5, the median is 77.5, and the sample standard deviation is 40.5. Check for skewness using Pearson’s skewness index.

Based on the PI, which is 0.148, the distribution for the data is not significantly skewed.

3(X median ) PI  s 3(79.5 77.5)  40.5  0.148

CH6: The Normal Distribution Santorico - Page 207 Example: Consider a data set with a mean of 126.5, a median of 144 and a standard deviation of 38.65, find the PI and interpret it.

PI=

CH6: The Normal Distribution Santorico - Page 208 Section 6-3: The

If we took numerous random samples and calculated the sample means, what distribution would the sample means have? This distribution is known as the distribution of the sample means.

Sampling distribution of sample means – the distribution of the sample means calculated from all possible random samples of size n from a population.

CH6: The Normal Distribution Santorico - Page 209 If the samples are randomly selected, the sample means will be somewhat different from the population mean . These differences are due to sampling error.

Sampling error - the difference between the sample measure  and the corresponding population measure due to the fact that the sample is not a perfect representation of the population.

Let’s look at a simulation through the (CLT) available at http://www.socr.ucla.edu/htmls/SOCR_Experiments.html.

What do you notice about the sampling distribution for the mean?

CH6: The Normal Distribution Santorico - Page 210 Properties of the Distribution of Sample Means

When all possible samples of a specific size are selected with replacement from a population, the distribution of the sample means for a variable has three important properties:

1. The mean of the sample means will be the same as the population mean and is given by

  . X

2. The standard deviation of the sample means (known as the of the mean) is given by

    . X n

Example: See pg 331 and 332 for a well-worked out example.

CH6: The Normal Distribution  Santorico - Page 211 The Central Limit Theorem

Reader’s Digest Version: The distribution of the sample mean gets closer and closer to a normal distribution as the sample size increases. This distribution has a mean    and a standard X  deviation   . X n

This result applies no matter what the shape of the probability distribution from which the samples are taken.



CH6: The Normal Distribution Santorico - Page 212 Some Important Points to Remember About the Central Limit Theorem:

1. If the distribution of the population is normal, then the distribution of the sample means will ALWAYS be normal.

2. If the sample size is large enough (typically n  30), then the distribution of the sample means will be approximately normal. The larger the sample, the better the approximation will be. 

CH6: The Normal Distribution Santorico - Page 213 Why is the Central Limit Theorem so useful?

Even when we don’t know the distribution of the population, if the sample size is sufficiently large, then we can use the properties of the normal distribution to make statistical inference about the population mean! So essentially, if the sample size is large enough we can make statistical inference about the population mean even if we don’t know anything else about the population!

If the population size is really large, then the results about the sampling distribution of the mean are approximately correct even if sampling without replacement.

CH6: The Normal Distribution Santorico - Page 214 When the sample size is sufficiently large (30+), the central limit theorem can be used to answer questions about sample means in the same way that a normal distribution can be used to answer questions about individual data. However, our z- score formula changes slightly to

x  z   / n since the standard deviation of x is  / n.

CH6: The Normal Distribution Santorico - Page 215   Example: The average yearly cost per household of owning a dog is $186.80 with a standard deviation of $32.00. Suppose that we randomly select 50 households that own a dog. What is the probability that the sample mean for these 50 households is less than $175.00?

We want to know Px( 175) and have =186.80, =32, n=50. To use our standard normal distribution, we need to convert to a Z- score:

x 175  175  186.80 P( x 175)  P   P Z  /nn / 32 / 50

PZ( -2.61) = 0.0045

CH6: The Normal Distribution Santorico - Page 216 What is the probability that the sample mean for these 50 households is between $195 and $200?

What is the probability that the sample mean for these 50 households is between $180 and $190?

CH6: The Normal Distribution Santorico - Page 217 Tips:

1. Look for keywords about whether the data are normally distributed or the sample size is relatively large. 2. Note whether you are looking for a probability or a percentile. 3. The formula z  (x )/ is used to gain information about an individual data value when the variable is normally distributed. x  4. The formula z  is used to gain information when  / n applying the central limit theorem about a sample mean when the population is normally distributed or when the sample size is 30 or more.. 

CH6: The Normal Distribution Santorico - Page 218